Journal of Economic Literature 2010, 48:4, 9871004 http:www.aeaweb.org/articles.php?doi=10.1257/jel.48.4.

987

WhyIsntMexicoRich?
GordonH.Hanson*
Over the last three decades, Mexico has aggressively reformed its economy, opening to foreign trade and investment, achieving fiscal discipline, and privatizing state owned enterprises. Despite these efforts, the countrys economic growth has been lackluster, trailing that of many other developing nations. In this paper, I review arguments for why Mexico hasnt sustained higher rates of economic growth. The most prominent suggest that some combination of poorly functioning credit markets, distortions in the supply of nontraded inputs, and perverse incentives for informality creates a drag on productivity growth. These are factors internal to Mexico. One possible external factor is that the country has the bad luck of exporting goods that China sells, rather than goods that China buys. I assess evidence from recent literature on these arguments and suggest directions for future research. (JEL E23,E65,F14,O10,O20,O47)

1. Introduction n1994,MexicojoinedtheOrganisationfor EconomicCooperationandDevelopment (OECD). Its admission, at the time unusual foraclubofprimarilyrichnations,wasrecognitionfromOECDmembersthatthecountry wasontheroadtosuccess.Afterasovereign default in 1982, which precipitated a currencycollapseandsharpcontractioninGDP, Mexicowasforcedtomakesignificantchanges initseconomy.Presagingreformeffortselsewhere in Latin America and a decade later in Eastern Europe, the country proceeded aggressivelytoreducetheroleofthestatein theeconomyandtoembraceglobalmarkets. Mexicos approach was nothing if not ambitious.Afteritsdefault,itfacedadifficult
*Hanson:UniversityofCalifornia,SanDiegoandNBER. ThankstoDanielChiquiar,EduardoEngel,RogerGordon, andSantiagoLevyforhelpfulcommentsandsuggestions.

macroeconomic stabilization, from which it emerged with an independent central bankandmoredevelopedfinancialmarkets (PedroAspe1993).Concomitantly,thecountryliberalizedforeigntradeandinvestment, firstunilaterally,byaccedingtotheGeneral Agreement on Trade and Tariffs, and later regionally,throughtheNorthAmericanFree Trade Agreement (NAFTA). In the span of afewyears,itprivatizedadisparatecollectionofnearly1,000state-ownedenterprises, including the communications behemoth, Telfonos de Mxico (Telmex), and the countrysbanks,whichhadbeennationalized inafailedefforttoimpedecapitaloutflows leading up to the 1982 crisis (Nora Lustig 1998).By1994,theyearNAFTAwasimplemented, Mexicos transformation seemed complete.Undertheguidanceofthreesuccessive technocratic governments, in which PhDeconomiststrainedatU.S.universities managed economic policy, the country had

987

988

Journal of Economic Literature, Vol. XLVIII (December 2010) (GordonH.Hanson2010).Thoughtherecent globalrecessionbroughtaseverecontraction in world imports in 2008 and 2009, robust tradegrowthhasnowresumed. In the 2000s trade boom, Mexico again seemed to miss out. Between 2001 and 2008,whiletheothercomparisoncountries managedannualgrowthratesof2.3percent or higher, Mexico was stuck at 1.3 percent (table1).Afterthreedecadesofsluggishness, the countrys growth record looks idiosyncratic, reflecting deep structural impedimentsratherthanshorttermproblems,such as real exchange rate overvaluation, which appearedresponsibleforthecountrysweak performance in the early 1990s (Rudiger Dornbusch and Alejandro Werner 1994). RecentliteratureprovidesahostofexplanationsforMexicoslaggardstatus.1Themost prominent suggest that some combination ofpoorlyfunctioningcreditmarkets,distortions in the supply of nontraded inputs, or perverseincentivesforinformalitycreatesa dragonproductivitygrowth.ThesearefactorsinternaltoMexico.OnepossibleexternalfactoristhatMexicohasthebadluckof exportinggoodsthatChinasells,ratherthan goodsthatChinabuys,meaningthatChinas expansion has put downward pressure on Mexicos terms of trade. Clearly, none of thesemechanismsisuniquetoMexico.For one or more of them to be able to explain Mexicosgrowthrecord,theymustafflictthe countryinaparticularlyseveremanner. Before examining the arguments for Mexicos stagnation, it is worth reviewing common explanations for slow growth that
1 For more in depth summaries of Mexicos growth performanceandpossibleexplanations,seeLustig(1998), which focuses on macroeconomic stabilization and the absence of growth in the immediate aftermath of economicreforms;thevolumesinSantiagoLevyandMichael Walton(2009),whichaddressthepoliticaleconomyofeconomicgrowthinMexico;andDanielChiquiarandManuel Ramos-Francia(2009)andJavierAriasetal.(2010),which examine the microeconomic foundations of Mexicos low productivitygrowth.

becomeaprimeexemplaroftheWashington Consensus (Pedro-Pablo Kuczynski and JohnWilliamson2003). Allthatwasmissingwaseconomicgrowth. Alas, the growth never came. Mexico succeeded in lowering inflation, maintaining fiscal discipline, reducing its external debt burden, and increasing trade as a share of GDP. But its growth has been lackluster. Figure1plotslogpercapitaGDPinMexico againstasetofcomparisoncountriesroughly similar in population and initial income. Within Latin America, Mexico has kept pacewithArgentina,butnotChileor,more recently, Brazil(panel A). Its growth iswell belowSoutheastAsia(panelB)andEastern and Central Europe (panel C). Between 1985 and 2008, Mexico managed an annual average growth rate in per capita GDP of just 1.1 percent, lower than all of the comparisoncountriesexceptVenezuela(0.8percent).Giventhevigorofitsreforms,itishard nottosensethatMexicohasunderachieved (Guillermo Ortiz 2003). Managing to beat only Venezuela, with its ongoing domestic economicandpoliticalturmoil,ishardlythe futurethecountryhadenvisioned.In2006,it wasinpartdisillusionmentwiththeeconomy that brought Mexico within a hairs breadth of electing a populist president who campaignedonreversingearlierliberalizations. Inthispaper,Ireviewargumentsforwhy Mexico hasnt sustained higher rates of economicgrowth.Muchhasbeenmadeofdevelopment disappointments in lower income countries(e.g.,WilliamEasterly2002;Joseph E. Stiglitz 2003). The 1990s, during which LatinAmericasufferedfromcontinuedmacroeconomic instability and a severe financialcrisisrockedAsia,appearedtobeliethe promiseofmarket-orientedeconomicreform. However,inthe2000ssuchconcernsreceded, astheexpansionofChinaandIndia,among other emerging nations, created a surge in global trade and contributed to broad-based incomegrowthinexport-orientedeconomies

Hanson: Why Isnt Mexico Rich?

Panel A. Latin America
1
Mexico Brazil Venezuela Argentina Chile

989

Log per capita GDP (1980 = 0)

0.5

0.5 1980 1985 1990 1995 2000 2005

Year Panel B. Southeast Asia

1.5
Mexico Malaysia Thailand Indonesia Philippines

Log per capita GDP (1980 = 0)

0.5

0.5 1980 1985 1990 1995 2000 2005

Year Figure 1:EconomicGrowthinComparisonCountries (continued)

990

Journal of Economic Literature, Vol. XLVIII (December 2010)

Panel C. Eastern and Central Europe

0.8
Mexico Hungary Turkey Bulgaria Romania

Log per capita GDP (1980 = 0)

0.6

0.4

0.2

0.2 1980 1985 1990 1995 2000 2005

Year Figure 1:EconomicGrowthinComparisonCountries(continued)

are insufficient to explain the Mexican case. BecausesomecountriesinLatinAmericahave donewellinthelastdecade,Mexicosperformancedoesnotappeartobesolelyattributable to the regionwide institutional deficiencies that are often blamed for the hemispheres slow development. These include a legacy ofSpanishandPortuguesecolonialism,high levelsofwealthinequality,andfactorendowments that facilitated the early emergence of a landed elite (e.g., Kenneth L. Sokoloff and Stanley L. Engerman 2000; Sebastian Edwards, Gerardo Esquivel, and Graciela Marquez 2007; Edwards 2009). The gap betweenaverageincomeintheUnitedStates andLatinAmericafirstmanifesteditselfinthe eighteenthcenturyandlaterwidenedinthe nineteenthandtwentiethcenturies.Between 1950 and 2001, there was zero convergence betweenLatinAmericaandtheUnitedStates in per capita GDP. Latin Americas decline

relativetoAsiaandEurope,whichduringthe secondhalfofthetwentiethcenturydidconvergetowardU.S.incomelevels,wasdueprimarilytototalfactorproductivity(TFP).The regionhaslongbeenplaguedbydisappointinglylowproductivitygrowth(HaroldL.Cole etal.2005).LatinAmericascommonhistory is surely important for understanding many aspectsofitseconomicdevelopment.Yet,in recentdecades,theregionhasnotmovedin lockstep.Chilehasperformeddecentlysince thelate1980sandBrazil,Colombia,andPeru have since 2000. Mexico, despite its market friendly reforms, has not joined the group, withitsperformanceclosertoArgentinaand Venezuela,theregionsblacksheepinterms ofeconomicpolicy.Itsstorymust,then,have plotlinesdistinctfromotherLatinnations. In low income countries, Paul Collier (2007) identifies common pitfalls that contribute to poverty traps. None apply to

Hanson: Why Isnt Mexico Rich?

991

RateofEconomicGrowth Averageannualchange logpercapitaGDP Country Argentina Brazil Bulgaria Chile Colombia CzechRepublic Hungary Indonesia Malaysia Mexico Peru Philippines Poland Romania Thailand Turkey Venezuela 19852008 0.016 0.014 0.022 0.042 0.019 0.020 0.018 0.035 0.035 0.011 0.015 0.013 0.039 0.012 0.045 0.027 0.008 20012008 0.032 0.023 0.062 0.031 0.031 0.040 0.035 0.038 0.031 0.013 0.044 0.028 0.042 0.068 0.037 0.033 0.027

TABLE1

Mexico, itself a middle income nation. For thelasteightyyears,Mexicohasavoidedmilitaryandpoliticalconflict,whichenveloped the country in the early twentieth century. Indeed,afterthe198294periodofintensive economic reform, the country ended seven decadesofsinglepartyruleandtransitioned to a functioning if combative democracy. Therehasbeenarecentupsurgeinviolence associatedwithdrugtrafficking,butintense conflictdidnotemergeuntil2006,toolateto account for Mexicos sluggishness.2 Despite
2 The growth of the drug trade may distort GDP figuresforMexico.Itislikelythatvalueaddedincultivating and exporting marijuana and opium, manufacturing and exporting methamphetamine, and distributing cocaine fromSouthAmericatotheUnitedStatesisnotfullyrepresentedinthecountrysnationalincomeaccounts.

being an oil exporter, the resource curse does not seem to have hit Mexico. Oil as a shareofexportspeakedin1982at77percent andhasdeclinedsteadilysince,fallingbelow 40percentby1988andbelow15percentby 1993.In2009,oilaccountedforjust13percentofMexicostotalexports.Overthelast twodecades,Mexicostradegrowthhasbeen heavilyconcentratedinmanufacturing(Juan Carlos Moreno-Brid, Jesus Santamaria, and Juan Carlos Rivas Valdivia 2005). If anything,Mexicohasinvestedtoolittleinpetroleum exploration and production, partly as a consequence of constitutional limitations on foreign investment in the sector (Osmel ManzanoandFranciscoMonaldi2008).And Mexicocannotbesaidtosufferfromresidinginapoorneighborhood.NAFTAaffords

992

Journal of Economic Literature, Vol. XLVIII (December 2010)

DomesticCredittoPrivateSector(percentofGDP) Country Argentina Brazil Colombia CzechRepublic Hungary Indonesia Malaysia Mexico Peru Philippines Poland Romania Thailand Turkey Venezuela 19912000 20.16 56.42 32.72 65.72 27.82 45.73 163.44 25.62 19.13 42.09 21.73 9.33 127.55 19.90 16.93 20012008 13.79 36.89 27.03 39.32 49.22 24.24 130.94 17.98 21.16 33.58 32.80 20.95 103.01 21.90 14.63

TABLE2

Notes: Domesticcredittoprivatesectorreferstofinancialresourcesprovidedtotheprivatesector,suchas throughloans,purchasesofnonequitysecurities,tradecreditsandotheraccountsreceivable.

Source:WorldDevelopmentIndicators.

thecountryaccesstotwonearbylarge,rich markets. In 2009, Mexicos exports to the United States as a share of its total foreign shipments were 81 percent. Mexicos problems,itseems,lieelsewhere. 2. Faulty Provision of Credit Theprovisionofcreditiscentraltotheprocess of economic development (Raghuram G. Rajan and Luigi Zingales 1998). Without mechanisms that move savings from lenders to borrowers, a country has poor prospects for taking advantage of productive investment opportunities. Mexico stands out for channelinglowlevelsofprivatecredittofirms or households. Over the period 200108, domestic credit to the private sector as a shareofGDPaveraged18percentinMexico,

lower than all comparison countries except ArgentinaandVenezuela(table2).Agrowing body of literature cites the weakness of Mexicoscreditmarketsasanimportantfactor behindthecountryslowproductivitygrowth. Poorly developed financial markets have plaguedMexicosinceitsindependencefrom Spain(StephenH.Haber1991).Oneproblem has been risk of expropriation. In the last forty years, the government has expropriated the assets of private banks twice (Haber, 2005). The first episode was a de factonationalizationthatoccurredinthemid 1970s, when the government increased the reservesbankswererequiredtoholdinthe centralbanktoover40percentandpaidan interestrateonthesedepositsthattrailedthe rate of inflation. The second expropriation wasanexplicitbanknationalizationin1982,

Hanson: Why Isnt Mexico Rich? whichoccurredinthemidstofMexicossovereign default. Another problem has been pooroversightofbanklending.Shortlyafter Mexicos banks were privatized in 1991, creditprovisionspiked.Haber(2005)argues that in Mexicos haste to privatize its banks itfailedtoestablishmechanismsthatwould impede the abuse of deposit insurance or lending to bank directors or other insiders. Consequently,bankcreditexpandedrapidly, financedinpartbyinterbanklending,which thegovernmentinsured.Rapidlyaccumulatingnonperformingloansandthe1994peso collapsecontributedtowidespreadbankfailures.3Thegovernmentwaslaterobligatedto recapitalizethebankingsystem. Sincethe199495bankingcrisis,Mexico has overhauled its financial regulations. It has modernized accounting standards, permitted foreign ownership of banks, set reserves according to the riskiness of bank loan portfolios, modified deposit insurance, and sought to prevent insider lending. While bankruptcy provisions were also modified, Haber (2005) suggests that these reformsprimarilyaffectedeasilycollateralizable loans, such as for autos or homes, but did little to spur commercial lending. As a result, the difficulty lenders have in seizing assets from borrowers continues to hinder credit provision. In the 2004 World Bank DoingBusinessIndicators,Mexicoscoreda 0outof4inprotectionsprovidedtocreditors, putting it among just eleven countries (outof132)inthelowestcategory.Intable 2, one sees that despite Mexicos reforms domesticcredittotheprivatesectoractually declinedfromthe1990stothe2000s,withits

993

performance relative to comparison countriesshowingnoimprovement.4 Turning to how credit markets affect growth,RaphaelBergoeingetal.(2002)suggest that Mexicos bankruptcy laws account for Mexicos poor performance relative to Chile. Beginning in the 1970s, both countries suffered government debt crises and boutsofhighinflationwhichtheyaddressed by imposing fiscal austerity and monetary restraint. Each liberalized trade and privatizedstateownedcompanies.Subsequently, Chileenjoyedhighproductivityandincome growthwhileMexicodidnot.Provisionsfor seizingassetsofborrowersindefaultarean important remaining difference between the two nations. Chiles laws are relatively favorable toward lenders. Through process ofelimination,Bergoeingetal.(2002)concludethatthedifferencesinbankruptcylaws are the most likely explanation for Chiles superior performance. While differences in bankruptcy provisions are certainly a plausible explanation for the two countries divergence, the evidence is circumstantial. Bergoeing et al. (2002) make a convincing casethatdifferencesinfiscalortradepolicy cannot account for Chiles relative success, but a host of other factors, including those mentioned below, remain outside consideration. Additionally, Bergoeing et al. (2002) argue that Chile and Mexico privatized state-owned companies to a comparable degree.Butotherliterature(e.g.,Levyand Walton2009)suggeststhatitishowMexico
4ChiquiarandRamos-Francia(2009)andHaber(2009) suggestthatwhilethesereformsdidnotappreciablylower interestratestoborrowerstheydidleadtogreaterconcentrationinthebankingsectorandtohigherprofitmargins (bothinabsolutetermsandrelativetoacomparisongroup of twelve other countries, among which only Brazil and Peruhadhigherbankprofitability).Mexicosgovernment competitioncommission,outofconcernthatthecountrys bankingsystemisimperfectlycompetitive,hasruledthat Mexicosbanksshouldmaketheirservicesmoretransparentandreducecoststoconsumersofswitchingbanks.It hasalsomovedtolowerbankentrybarriers.

3 Aaron Tornell, Frank Westermann, and Lorenza Martinez (2004) suggest that because the credit crunch wasaccompaniedbyasignificantrealexchangeratedepreciationthenon-tradedsectorwashitparticularlyhard,with thecontractioninnontradablesultimatelycreatingbottlenecksinthesupplyofinputsthatinhibitedexportgrowth.

994

Journal of Economic Literature, Vol. XLVIII (December 2010) that higher exemptions for personal bankruptcy in U.S. states operate as a form of wealth insurance, which increases the level of entrepreneurship, even as high risk borrowersarecutofffromloans.5 Howimportant,then,areMexicoscredit problems in explaining its growth performance?Basedonasurveyof4,000firmsin fifty-four countries during the late 1990s, Thorsten Beck, Asli Demirguc-Kunt, and VojislavMaksimovic(2005)reportthatthe fraction of firms saying they face severe obstacles in obtaining finance is highest in Mexico. Their work also finds that firm salesgrowthislowerincountriesthathave greater obstacles in obtaining financing, measured in terms of collateral requirements, bank paperwork, or interest rates. These findings, while intriguing, represent indirect evidence on economic growth, as theyrelatetofirmsalesandnotproductivity and do not account for why financial policiesvaryacrosscountries. Onecouldimaginetwoapproachestoestimatetheimpactofbankruptcyprovisionson productivity growth in Mexico. One would beacrosscountryapproach,similartoBeck, Demirguc-Kunt,andMaksimovic(2005)but extendedoveralongertimeperiod,inwhich one regressed firm or industry growth in productivityonmeasuresofcreditorprotection, exploiting cross country and cross time variationinbankruptcyprovisions.Theobviousproblemwithsuchregressionsisthatin orderforthemtobewellidentifiedonewould havetofindsourcesofexogenousvariationin countries financial provisions, which could
5 For firms in the informal sector, the incentives for entrepreneurship from generous bankruptcy provisions may be undermined by other aspects of the tax code. Informalentrepreneurs,byvirtueoftheirtaxstatus,reap notaxsavingsfromlossestoincomeorcapital.However, theymayonlybeabletorealizesubstantialgainstoincome orcapitalbyshiftingintotheformalsector,whichbrings withitanincreaseintaxliabilities.Theasymmetrictreatmentofgainsandlossesmaybeadisincentivetoinnovationamonginformalenterprises.

privatizeditscompaniesthatmatters,inparticular the countrys failure to prevent the formationofprivatemonopolies. For credit markets to be a factor in Mexicoslowgrowthrates,theymustaffect productivity.Togaugetheextentofbarriers to investment in Mexico, David McKenzie andChristopherWoodruff(2008)conducted anexperimentinwhichtheyrandomizedthe provisionofcashgrantstosmallretailfirms in Mexico, the results from which allow themtoestimatethemarginalrateofreturn toinvestment.Theyfindthatratesofreturn to capital are 20 percent to 32 percent per month,whichwerethreetofivetimesmarketinterestratesduringthesampleperiod. Rates of return are highest among firms that report being financially constrained (but,surprisingly,alsoremainabovemarket ratesforfirmsthatdonotreportbeingconstrained).Theirresultssuggestthatthereare imperfectionsinMexicosfinancialmarkets, which impede the flow of credit to profitable undertakings, particularly for small scale entrepreneurs. Other evidence, some experimentalandsomenot,suggeststhatin Ghana,India,Kenya,andSriLankaratesof returntocapitalinsmallscaleenterprisesare alsowellabovemarketrates(seediscussion inMcKenzieandWoodruff2008),implying thatMexicossituationisnotunique. Could weak protections for creditors be responsible for limited access to credit in Mexico?IntheUnitedStates,ReintGropp, John Karl Scholz, and Michelle J. White (1997)findthat,instatesthatprovidehigher personal exemptions for debtors in bankruptcy,whichlimitstheassetsthatcreditors can seize, a lower fraction of credit is held by low-asset households, suggesting that weak creditor protections divert lending away from relatively poor borrowers. Yet, divertingcreditawayfromthepoor,whilea concernforpovertyalleviation,doesnotnecessarilytranslateintodiminishedproductivity growth. Wei Fan and White (2003) find

Hanson: Why Isnt Mexico Rich? bedifficult.Asecondapproachwouldbefor Mexicoitselftoexperimentwithchangingits bankruptcy provisions, perhaps by varying the timing of adjustments in protections to creditors across states within the country. Onecouldthenestimatehowfirm,industry, or regional outcomes change in response to thepolicychange.Suchanexperimentwould clearlybeasignificantundertaking,butgiven the low apparent return to Mexico from its extensive financial reforms to date the benefitsfromidentifyingwhichfeaturesofcredit marketscontinuetoinhibitgrowthmayyield commensuratelylargereturns. 3. Social Policy and Informality The size of the informal sector in Latin Americaisfrequentlycitedasadragoneconomic development. Aside from Hernando deSotos(1989)so-calledromanticperspective, in which informal firms are a vibrant andunderappreciatedpartoftheproductive economy,manycommentatorsseeinformalityastheresultoffirmsremainingsmallin order to avoid onerous government regulations (Rafael La Porta and Andrei Shleifer 2008). Informality keeps firms in existence thatwouldbeforcedtoexit,eitherbecause ofpoormanagementoroutdatedtechnology, if they had to compete for inputs with formalsectorfirmsonalevelplayingfield.One consequenceofinformalityis,therefore,the survival of small, unproductive enterprises. Thedispersioninproductivityacrossplants inMexicoismuchhigherthanintheUnited States. In 2004, 36 percent of manufacturingemploymentinMexicowasinestablishmentswithfewerthanfiftyworkersand22 percent was in establishments with fewer than ten workers. Enterprises with fewer than ten employees tend to have very low TFP(CarmenPages2010),with91percent of these plants having productivity levels belowtheirindustryaverage.Themajorityof thesemicro-enterprisesareinformal.

995

Chang-Tai Hsieh and Peter J. Klenow (2009) use micro data on Mexican nonagricultural establishments to examine how aggregateTFPinthecountrywouldchange were productivity equalized across establishments by moving factors toward more productive plants, while accounting for endogenous changes in productivity at the plant level. Based on data for 1999 and 2004,theyestimatethatsuchareallocation would raise TFP in Mexico by 100 percent inthemanufacturingsectorandby250300 percent in nonmanufacturing. The aggregate productivity gain from factor reallocationforMexicois,interestingly,comparable to that for China but 1.5 to 2 times larger thanforArgentina,Bolivia,Chile,Colombia, Ecuador,orUruguay. For these counterfactual exercises to be meaningful,theremustbedistortionsinthe economy that prevent factor reallocation fromoccurringnaturally.HsiehandKlenow find that TFP is higher in firms that make payments into Mexicos social security system,whichisLevys(2008)preferreddefinitionofformality,suggestingthatevasionof payroll taxes is among the distortions contributingtoproductivitydispersion.Formal sector firms make contributions to social security equal to 18 percent of the total wagebill. Hsieh and Klenow estimate that reallocating factors to establishments that pay into social security would raise aggregateproductivityby19percentinmanufacturing,oronefifthofthetotalproductivity gainfromfactorreallocation,implyinginformalityisanimportantsourceofcross-plant differencesinTFP. Levy (2008) makes an explicit argument for how informality affects productivity in Mexico. The starting point for his critique is Mexicos dual system of social insurance. One system governs the formal economy, in whichworkersandemployersaresubjectto obligatorypayrolltaxesthatcoversocialsecurity benefits, including retirement pensions,

996

Journal of Economic Literature, Vol. XLVIII (December 2010) alowerlevelofTFPbutdoesitsuppressTFP growth?InMexicoscase,onewouldwantto knowwhetherraisingthebenefitstoinformal workers through social protection programs (financed by consumption taxes) lowers the growth rate. Research has yet to provide a definitiveanswertothisquestion. What research has examined is whether the recent expansions of social protection programs available to informal workers have reallocated employment toward the informal sector. The aggregate data show noupwardtrendininformality(Ariasetal. 2010).Between1995and2009,thefraction of workers employed outside of the social securitysysteminMexicoactuallydeclined, from66percentto62percent,thoughinformalitydidriseamongurbanworkers.Among the different categories of informal workers,onlytheshareofillegalsalariedlaborin employment has increased over time. The employment share of self-employed workershasbeenstableandtheshareforworkers paidoncommissionhasfallen. Inmicro-levelanalysisofwhetherMexicos dualsystemofsocialinsurancehasincreased informality, Rodrigo Barros (2009) exploits regional variation in the timing of the role outofSeguroPopular,alargehealthinsurance program available to workers who are not covered by social security which is the centerpieceofMexicosnewsystemofsocial protection.Theprogramamountedtoasubstantial increase in the quality and quantity of health insurance to qualifying individuals,whoaccountforroughlyhalfthepopulation, forcing the government to phase in the program gradually. The timing of the roll out favored small states and states not in control of the main opposition political party.Barrosfindsthatinrepeatedcrosssection data the probability a household head worksintheinformalsectorisuncorrelated with the government projected enrollment intensityofSeguroPopular(i.e.,theplanned enrollment in a state in a given year based

health care, disability and life insurance, childcare,andhousingloans.Formalsector employmentisalsosubjecttoregulationsthat govern firing and severance pay. A second systemofsocialinsurancecoverstheworkers thatLevydefinesasinformal,includingillegallyemployedsalariedworkers(i.e.,workers thatbylawshouldbeenrolledinsocialsecuritybutarenot),workerspaidoncommission, theselfemployed,andunpaidworkers.These individualshavetheoptionofparticipatingin socialprotectionprograms,subjecttoanominal contribution, which provide health care, retirementpensions,housingsubsidies,child care,andlifeinsurance.Importantly,theparticipation of informal workers in social protection programs is voluntary and a la carte, allowingindividualstoselectintothespecific programsthatsuittheirneeds.Formalworkers,whosecontributionstosocialsecurityare mandatory,havenosuchoption.Further,the monetary contributions of informal workers donotcoverthefullcostofsocialprotection programs,withthedifferencemetinpartby consumptiontaxes. Intheory,thetaxonformalsectoremployment and the subsidy to informal sector employmentresultsinarelativelylow(high) marginalproductivityoflaborintheinformal (formal) sector, creating static inefficiency and depressing aggregate TFP. Empirical evidence, as seen, is consistent with such outcomes.Additionally,informalitymayhave negative dynamic effects on productivity, and therefore growth, if keeping wages of formal establishments artificially high reallocatesinvestmenttowardlowerproductivity informalfirms.InMexicoitappearsthatthe returnstolabormarketexperiencearelower ininformalthaninformalemployment,which may indicate that informality suppresses the accumulation of human capital (Arias et al. 2010). While the theoretical linkages between informality and growth are clear, making the empirical connection is more problematic.Informalityappearstoresultin

Hanson: Why Isnt Mexico Rich? ontherolloutstrategy),meaningthatindividuals in states exposed to earlier roll out of Seguro Popular were not more likely to be employed informally. Because program intensityisnotrandom,theresultsarehard to interpret. The estimation strategy would underestimatetheeffectofSeguroPopular oninformalityiftheplannedrolloutofthe programwasmorerapidinstatessubjectto shocksthatdisfavorformality. Melissa A. Knox (2008) addresses similar questions using panel data on 800 families, exploiting temporal variation in which families were eligible for Seguro Popular (thus allowing her to control for household fixed effectsininformalstatus).Becomingeligible fortheprogramincreaseslaborforceparticipation,whichimpliesanincreaseinemploymentintheinformalsectorgiventhateligible households are those that do not qualify for socialsecurity.LauraJurez(2008)focuseson therolloutofafreehealthcareprogramin MexicoCity,findingthatforfemales(insalaried employment with no more than twelve yearsofeducation)expandedhealthbenefits reducesthelikelihoodofbeingemployedformally(i.e.,beingcoveredbysocialsecurity). Though the literature is still young, there is some evidence that expanded health insuranceforinformalworkersincreasesemploymentintheinformalsector. ThelevelofTFPininformalestablishments in Mexico is low relative to formal plants. Mexicos social protection programs may be raisingtheincentiveforinformality,possibly hindering growth. However, the newness of major financing for these programs suggests theycannotaccountforMexicospoorgrowth performance prior to the late 1990s. Social protection programs for the poor in Mexico haveaccommodatedinformality,therebycontributingtoitspersistence,butcannotexplain the origins of informality in Mexico or why thegrowtheffectsofinformalityinthecountry would be worse than elsewhere in Latin America.

997

4. Too Little Regulation (or Too Much) In comparing production costs across countries,Mexicostandsoutforhavinghigh pricesforelectricity,highpricesfortelecommunications services, expensive and spotty internet services, and a scarcity of skilled labor. In each of these input markets, specific interest groups are blamed either for impeding the government from enforcing anti-monopolyregulationsorforconvincing thegovernmenttoregulatemarketsinaway that preserves their capacity to earn rents (LevyandWalton2009).WhileIdiscussjust three input markets, these arguments have been applied to other sectors of Mexicos economy,aswell(seeIsabelGuerrero,Luis Felipe Lpez-Calva, and Walton 2009). While unfavorable conditions in Mexicos inputmarketsarewelldocumented,itisless clearwhethertheycanaccountforMexicos poorrelativegrowth. ThemostmalignedindustrialistinMexico is perhaps Carlos Slim, the worlds richest personandthelargestshareholderinTelmex, Mexicos telecommunications giant. Slim is blamed for using his acquisition of Telmex from the government first to establish and later to protect monopoly power in markets for landline telephone services, mobile telephony, and internet access. In data for 200608, Mexico compares poorly to other countriesinthecostsofitstelecomservices. Amongagroupofdevelopinganddeveloped nations,6 it has the highest levels of singlefirmconcentrationinfixedandmobiletelephone industries (with Telmex controlling over 90 percent in the former and over 70 percent in the latter), the highest costs of fixedlinebusinesstelephoneservices(which aretwoandhalftimesthelevelintheUnited
6 ThecomparisoncountriesareBrazil,Canada,Chile, China, Colombia, the Czech Republic, EU nations, Hungary,Japan,Korea,NewZealand,Poland,theSlovak Republic,Turkey,andtheUnitedStates.

998

Journal of Economic Literature, Vol. XLVIII (December 2010) Mexicomayhavelowergrowthrates,resultingeitherfromareducedsupplyofinnovation-producingscientistsandengineers(e.g., CharlesI.Jones1995)orfrommissingouton positive externalities in the accumulation of humancapital(e.g.,RobertE.Lucas1988). In2007,only50percentofMexicans1519 years old were enrolled in public or private educational institutions, slightly above levelsforTurkeybut20percentto42percent belowBrazil,Chile,Estonia,Hungary,Korea, Portugal, the Slovak Republic, and Slovenia (Arias et al. 2010). Mexicos poor educationperformancemaybeattributabletothe captureofitseducationsystembypowerful labor unions. Close to 90 percent of public educationexpenditureinMexicoisoncompensation of staff, leaving little left over for buildings, computers, or other educational infrastructure. However, other countries in Latin America also have unionized public education,whichraisesthequestionofwhy Mexicos would have consequences that are more deleterious for growth. Mexicos labor unionshavestronglyopposedlinkingteacher pay to student performance but this hardly makesthemunique. Beyond rates of tertiary education, Mexican students perform poorly in standardized tests relative to students in other nations at roughly similar income levels. Combining PISA scores for math, reading and science, Mexico does worse than all developing countries except Brazil, scoring well below Chile, the Czech Republic, Estonia, Hungary, Korea, Poland, Russia, and the Slovak Republic, both in absolute terms and relative to what one would predictbasedonexpenditureperstudentinthe country (Arias et al. 2010). Lant Pritchett and Martina Viarengo (2009) suggest that what may matter for a countrys economic developmentisnottheaverageperformance ofstudentsbutthenumberofhighachievers.Hereagain,Mexicosdoesnotcompare well.APISAscoreof625qualifiesastudent

States),andthelowestnumberofbroadband internet subscribers per capita (Chiquiar and Ramos-Francia 2009). While it is plausiblethatTelmexsdominantmarketposition allows it to earn supra-normal profits, it is unclearwhethersuchanoutcomemattersfor Mexicos growth rate. The argument would havetobethathighpricesfortelecomservices depresscapitalaccumulationorinnovation.A causallinkbetweenthepriceandavailability oftelecommunicationsservicesandeconomic growthhasyettobeestablished. Surprisingasitmayseemforacountrythat exportsoil,Mexicohasrelativelyhighprices for electricity. Since 2000, electricity prices per kilowatt hour in Mexico have exceeded thoseintheUnitedStatesby1.1to1.7times. Among comparison countries, Mexico has thehighestrateofenergyloss,measuredas energy produced but not paid for as a percentage of total energy handled (Chiquiar and Ramos-Francia 2009). Electricity productioninMexicoisthesoleprovenanceof state owned companies, which have historicallyhadstronglaborunions.Theproximate causesforMexicoshighcostsandlowquality in electricity generation include reliance on oilbasedproduction,lowlaborproductivity, and high wages for electricity workers, who are among the most highly paid workers in thecountryandwhoseaveragewageratesare overthreetimestheaverageforformalsector workers. Holding output constant, were electricity generation in Mexico to operate withthesamelaborproductivityasinChile, Mexico would have 62 percent fewer workersinthesector.Aswithtelecommunications services,highelectricitypricesraiseproduction costs relative to other countries. Again, however,itisunclearhowthepriceofelectricityaffectstherateofeconomicgrowth. An input market whose functioning does have well established effects on growth is thesupplyofskilledlabor.IfMexicospublic schoolsorlabormarketregulationssomehow reducetheincentivetoattaineducation,then

Hanson: Why Isnt Mexico Rich? as advanced. Only 0.3 percent of Mexican studentsattainascoreof625orhigher,relativeto18percentinKorea,9percentinthe SlovakRepublic,and2percentinThailand. If such high achievers are used intensively in developing or applying new technology, MexicomayfacesupplyconstraintsinR&D relativetoothercountries. Another factor potentially affecting Mexicossupplyofskilledlaborisemigration totheUnitedStates.Asof2008,therewere 11.4millionindividualsborninMexicoresidingintheUnitedStates,equalto9.5percent ofthetotalMexicobornpopulation(defined as the sum of Mexico born individuals in MexicoandtheUnitedStates).Emigrantsare overrepresented among the young. In 2000, 18percentofMexicobornmenaged16to25 wereresidingintheUnitedStates,compared to9percentof46to55yearolds.However, theskillprofileofemigrantsdiffersonlymodestlyfromindividualsthatremaininMexico. Chiquiar and Hanson (2005) compare the earnings of nonmigrants in Mexico with the earningsthatemigrantswouldearnwerethey toreturntothecountryandbepaidaccording to prevailing wage patterns. Emigrant men aredrawndisproportionatelyfromthemiddle of Mexicos wage distribution, whereas emigrantwomenaremodestlypositivelyselected. These results are based on 1990 and 2000 censuses for Mexico and the United States, inwhichonemightbeconcernedaboutthe undercount of illegal immigrants. Robert Kaestner and Ofer Malamud (2010) largely replicate Chiquiars and Hansons findings usingtheMexicanFamilyLifeSurvey,which tracksindividualsinMexicoovertime,including those that migrate to the United States, whetherlegallyorillegally. Itislikelythat,inMexico,emigrationand economicgrowtharejointlydetermined.Slow growthtodaymaycontributetotheexodusof labor,whichtomorrowinhibitsopportunities for growth. We know little about how emigration affects skill or capital ccumulation a

999

in Mexico. Higher emigration puts upward pressure on Mexican wages (Prachi Mishra 2007; Abdurrahman Aydemir, and George J. Borjas 2007), educing the incentive r for capital inflows. Since the majority of Mexican migrants enter the United States illegally (Hanson 2006), they may perceive weak incentives to continue their education beyond secondary school. Thus, emigration may be contributing to Mexicos observed low rates of school attendance among 15 to 19yearolds.InruralMexico,(McKenzieand Hillel Rapoport forthcoming) find that the likelihoodofcompletingsecondaryschoolor beginninghighschoolislowerforyouthsin householdsthathavesentmigrantsabroad. While we see that Mexico produces few highly educated students relative to other countries, the literature has yet to explain whyorquantifytheimplicationsofthisfact foreconomicgrowth.RodrigoGarca-Verd (2007) estimates that rising educational attainmentspecifically, the decline in the share of workers with no schooling or primary schooling and in the increase in the sharewithsecondaryortertiaryschooling accounts for 74 percent of Mexicos GDP growth over 198894 and 65 percent over 19952005.However,aggregateTFPgrowth wasnegativeinbothtimeperiods.Ifeducational attainment was rising, why was TFP declining?Thisquestioncannotbeanswered inastandardgrowthaccountingframework, in which TFP growth is taken as given. Obvious directions for future research are toexaminetheinteractionsbetweenschooling,academicperformance,emigration,and TFPlevelsandgrowthratesinMexico. 5. The Perils of Competing with China Inthe1980s,asMexicoembarkedupona strategyofexport-leddevelopmentitsought to emulate the successful industrializers of East Asia. Then President Carlos Salinas cited Korea, in particular, as a model for

1000

Journal of Economic Literature, Vol. XLVIII (December 2010) LatinAmerica.OtherlargeLatinnationstend toexporttheminerals,agriculturalgoods,and other commodities that China imports. In Mexico,asalreadymentioned,petroleumnow accountsforlessthan15percentofexports. Forbetterorworse,Mexicosfortunesinthe global economy are tied to manufacturing. Over the period 20002005, manufacturing accountedfor88percentofexportsinChina and83percentinMexico,butonly54percent inBrazil,34percentinColombia,20percent inPeru,and16percentinChile. Figure2showsthatovertheperiod1991 to2007,ChinasshareofU.S.manufacturing importsrosesteadily,whereasMexicosshare rose through the 1990s and then began to decline.Forbothcountries,thereisaninflection point in 2001, at which time Chinas increaseinU.S.marketshareacceleratedand Mexicochangedfromhavingarisingshareto a declining one. The significance of 2001 is thatthisyearmarksChinasaccessiontothe World Trade Organization, giving it most favorednationstatusamongWTOmembers. ItisplausiblethatChinasjoiningtheWTO helped it to capture foreign market share fromMexico(see,e.g.,IreneBrambilla,Amit K.Khandelwal,andPeterK.Schott2010on theapparelandtextileindustries). ThechallengeofChinasexportgrowthfor Mexicoisthatthetwocountriesspecializein similar goods. Chiquiar and Ramos-Francia (2009) calculate correlations between indices of revealed comparative advantage for Mexico and for other developing countries. Mexicosrankingofgoodsbyrevealedcomparative advantage is most strongly correlated with Hungary, followed in close order byThailand,thePhilippines,Korea,Turkey, Poland, and China. They find further that in the last decade Mexico has been losing comparative advantage vis--vis China in around 40 percent of its manufacturing exports (including electrical machinery, computers and lectronics, furniture, and e nonmetallic minerals), while maintaining a

Mexico, with hopes that the country would createitsownchaebolKoreaslargeindustrialgroupscharacterizedbyverticalintegrationinproduction,horizontaldifferentiation inoutput,andrapidexportgrowthtolead the country into the global economy. The realityturnedouttobefardifferent.Exports have grown dramatically in Mexico, rising from12percentofGDPin1982to28percent in 2008. However, it is not large, vertically integrated firms that export the bulk ofMexicosgoods.Instead,nearlyhalfofthe countrysmanufacturingexportsandover20 percentofitsmanufacturingvalueadded(as of 2006) are produced by maquiladoras, or exportassemblyplants,whichimportinputs from abroad, primarily the United States, assemble or process the inputs into final outputs,andthenexportthefinishedgoods, again primarily to the United States (Paul R.Bergin,RobertC.Feenstra,andHanson 2009). Hong Kong and Taiwan also began their industrial development with heavy reliance on export assembly. Their firms later graduated into original-equipment manufacture and own-brand production, which involves greater in-house fabrication of inputs and design of outputs (Michael J. Enright,EdithE.Scott,andDavidDodwell 1997). Mexico has not made such a transition.Whilethecountryhasprogressedfrom assemblingappareltoassemblingelectronics andautoparts,itremainsspecializedinthe labor-intensive processing of inputs for the U.S.economy. Mexicos failure to graduate from export assembly has left it exposed to competition fromChina,whoseabundanceinlowskilled laborhasgivenitastrongcomparativeadvantageinlabor-intensivemanufacturing.Bythe late1990s,exportprocessingplantsinChina, which are similar to Mexicos maquiladoras, accounted for over half of the countrys manufacturingexports(FeenstraandHanson 2005). For Mexico, Chinas growth is more threatening than it is for other countries in

Hanson: Why Isnt Mexico Rich?

1001

0.1
Mexico China

0.08

0.06

0.04

0.02 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008

Year Figure 2:ShareofU.S.ManufacturingImports

comparative advantage in another third of itsproducts(includingautomobilesandauto parts, industrial machinery, and beverages). Hanson and Raymond Robertson (2010) use results from a gravity model of trade to estimatethechangeindemandforMexicos exportsthatwouldhaveoccurredhadChinas export supply capacity remained constant over the period 1995 to 2005. For Mexico, nullifying the improvement in Chinas export capability would mean a 2 percent to 4 percent increase in the global demand for Mexicos exports, an effect that is larger than for any other manufacturing oriented developingeconomy.7HsiehandRalphOssa
7 The comparison countries are Hungary, Malaysia, Pakistan, the Philippines, Poland, Romania, Sri Lanka, Thailand,andTurkey.

(2010) take a more theoretical approach, introducing Ricardian productivity differences into a Marc J. Melitz (2003) model of trade, deriving omparative tatics for c s changes in wages and prices resulting from productivity changes in different countries, calibratingthemodeltodata,andthencalculating the counterfactual income levels that would have been obtained had productivity growth in China over 199394 to 200405beenzero.Notsurprisingly,slower productivity growth in China means higher realincomesinMexico.HadChinasproductivitybeenflat,Mexicoswelfarewouldhave been0.8percenthigher,alargereffectthan foranyothereconomythatHsiehandOssa consider. Asidefromeffectsonthelevelofincome in Mexico, could competition from China

1002

Journal of Economic Literature, Vol. XLVIII (December 2010) 6. Final Discussion Anydiscussionofgrowthanddevelopment inMexicoendsupresemblingaDiegoRivera mural,overstuffedwithhistoricalcharacters that collide in repeated and unexpected ways.Mexicosunderperformanceisoverdetermined.Thefaultyprovisionofcredit,persistence of informality, control of key input marketsbyelites,continuedineffectiveness of public education, and vulnerability to adverseexternalshockseachmayhavearole in explaining Mexicos development trajectoryoverthelastthreedecades.Wedonot yet know the relative importance of these factorsforthecountrysgrowthrecord. Perhaps what is most striking about Mexicosrecentexperienceisthatthecountry has tried so hard but achieved so little. The breadth and depth of reform in the country is astounding, yet Mexico does not have much to show for it. Going forward, onewouldhopethatfutureresearchwould focus on providing guidance to policymakers about which specific policy interventions may be effective. One can imagine individual research agendas to estimate the impactofchangesinbankruptcyprovisions, social insurance programs, anti-monopoly regulations, government regulation of electricitymarkets,orpubliceducationonindustrial or regional productivity. Suchresearch would undoubtedly be of use to Mexican policymakers.Butthereisadangerinfocusing exclusively on individual policies and ignoring their interactions. Even a cursory examinationofMexicoseconomicstructure suggeststhecountryisdeepintheworldof thesecondbest,meaningthatmodelsbased onasingledistortionmaybeapoorguideto howfuturereformwouldchangeeconomic outcomesinthecountry. An example from the discussion in this paper helps drive the point home. Bankruptcy provisions that are overly protectiveofborrowersinhibittheflowofcredit

explain Mexicos lackluster economic growth?Ifso,theargumentwouldgoasfollows.Sincethelate1980s,Chinahasbeen undergoing a transition from a rural economydominatedbystateownedenterprises toanurbanandmoremarket-orientedone. The transition is protracted, as it involves moving over a 150 million workers from thecountrysidetocities,importingforeign technology,openingChinasmarketstothe restoftheworld,andallowingprivatefirms to replace the state in production. Chinas size, high rate of growth, and increasing outward orientation mean that its emergence is surely changing international prices, improving the terms of trade for countriesthatproduceitsimportablesand deteriorating the terms of trade for countries that produce its exportables. Mexico fits squarely in the latter camp, whereas Argentina, Brazil, Chile, Colombia, and Peru fit in the former. Nevertheless, the downwardpressureonthepriceofMexicos export goods is unlikely to be permanent. ItwillabateonceChinacompletesitstransition or accumulates enough human and physical capital to graduate from the cone of diversification that Mexico now occupies and to move to a cone specializing in more skill and capital intensive products, aprocesswhichmayalreadybeunderway. The impact of China on Mexico, then, is notaboutlongrungrowthbutaprolonged transitionalperiodinwhichChinahasbeen realizingitslatentcomparativeadvantage. The main problem with the China hypothesis for Mexicos growth record is timing.Infigure2,itisapparentthatcompetition from China became most acute after2001,suggestingChinamayhavehad onlyamodestroleinMexicosslowgrowth inthe1980sandearly1990s.WhileChinas continued rise may not be good news for Mexico,itcannotbethesolefactorbehind Mexicos performance over the last three decades.

Hanson: Why Isnt Mexico Rich? toindividualswhoseobservablecharacteristics suggest they are poor credit risks. The high rate of return to investment in smallscaleenterprisessuggeststhatcreditmarket imperfections impede productive investments.Yet,thehighrateofreturntocapital exists in small-scale enterprises that other data suggest have very low TFP and whose survival depends in part on evading payroll taxesthatformalsectorfirmsdonotescape. If the distortions that cause informality are maintained, reforming bankruptcy provisions could be counterproductive in that it wouldshiftinvestmenttowardsthecountrys leastproductiveestablishments.Ifonewants toconsiderreformingbankruptcyprovisions in Mexico, one should also seek to understand how these provisions interact with socialinsuranceprogramsthatmaypromote informality. Research, then, would benefit from proceeding along two tracks, one that performsrigorousmicroeconomicanalysisof individual policy interventions and another thatuncoversthegeneralequilibriuminteractionsofthesepolicies.
References
Arias, Javier, Oliver Azuara, Pedro Bernal, James J. Heckman, and Cajeme Villarreal. 2010. Policies to Promote Growth and Economic Efficiency in Mexico.InstitutefortheStudyofLaborDiscussion Paper4740. Aspe,Pedro.1993.Economic Transformation the Mexican Way.CambridgeandLondon:MITPress. Aydemir, Abdurrahman, and George J. Borjas. 2007. Cross-Country Variation in the Impact of International Migration: Canada, Mexico, and the United States.Journal of the European Economic Association,5(4):663708. Barros, Rodrigo. 2009. Wealthier but Not Much Healthier:EffectsofaHealthInsuranceProgramfor thePoorinMexico.StanfordInstituteforEconomic PolicyResearchDiscussionPaper09-002. Beck, Thorsten, Asli Demirguc-Kunt, and Vojislav Maksimovic.2005.FinancialandLegalConstraints to Growth: Does Firm Size Matter? Journal of Finance,60(1):13777. Bergin, Paul R., Robert C. Feenstra, and Gordon H. Hanson. 2009. Offshoring and Volatility: Evidence from Mexicos Maquiladora Industry. American Economic Review,99(4):166471. Bergoeing, Raphael, Patrick J. Kehoe, Timothy J.

1003

Kehoe,andRaimundoSoto.2002.ADecadeLost andFound:MexicoandChileinthe1980s.Review of Economic Dynamics,5(1):166205. Brambilla, Irene, Amit K. Khandelwal, and Peter K. Schott.2010.ChinasExperienceundertheMultifiber Arrangement (MFA) and the Agreement on Textiles and Clothing (ATC). In Chinas Growing Role in World Trade, ed. Robert C. Feenstra and Shang-Jin Wei, 34587. Chicago and London: UniversityofChicagoPress. Chiquiar,Daniel,andGordonH.Hanson.2005.InternationalMigration,Self-Selection,andtheDistributionofWages:EvidencefromMexicoandtheUnited States.Journal of Political Economy,113(2):23981. Chiquiar, Daniel, and Manuel Ramos-Francia. 2009. CompetitivenessandGrowthoftheMexicanEconomy.BancodeMexicoWorkingPaper2009-11. Cole,HaroldL.,LeeE.Ohanian,AlvaroRiascos,and JamesA.Schmitz.2005.LatinAmericaintheRearviewMirror.Journal of Monetary Economics,52(1): 69107. Collier,Paul.2007.The Bottom Billion: Why the Poorest Countries Are Failing and What Can Be Done About It.OxfordandNewYork:OxfordUniversityPress. deSoto,Hernando.1989.The Other Path: The Invisible Revolution in the Third World.NewYork:Harper &Row. Dornbusch, Rudiger, and Alejandro Werner. 1994. Mexico: Stabilization, Reform, and No Growth. Brookings Papers on Economic Activity,1:25397. Easterly,William.2002.The Elusive Quest for Growth: Economists Adventures and Misadventures in the Tropics.CambridgeandLondon:MITPress. Edwards,Sebastian.2009.LatinAmericasDecline:A LongHistoricalView.NationalBureauofEconomic ResearchWorkingPaper15171. Edwards, Sebastian, Gerardo Esquivel, and Graciela Marquez,eds.2007.The Decline of Latin American Economies: Growth, Institutions, and Crises. ChicagoandLondon:UniversityofChicagoPress. Enright,MichaelJ.,EdithE.Scott,andDavidDodwell. 1997.The Hong Kong Advantage.OxfordandNew York:OxfordUniversityPress. Fan,Wei,andMichelleJ.White.2003.PersonalBankruptcy and the Level of Entrepreneurial Activity. Journal of Law and Economics,46(2):54367. Feenstra, Robert C., and Gordon H. Hanson. 2005. Ownership and Control in Outsourcing to China: EstimatingtheProperty-RightsTheoryoftheFirm. Quarterly Journal of Economics,120(2):72961. Garca-Verd,Rodrigo.2007.Demographics,Human Capital and Economic Growth in Mexico: 1950 2005.Unpublished. Gropp,Reint,JohnKarlScholz,andMichelleJ.White. 1997. Personal Bankruptcy and Credit Supply and Demand. Quarterly Journal of Economics, 112(1): 21751. Guerrero, Isabel, Luis Felipe Lpez-Calva, and MichaelWalton.2009.TheInequalityTrapandIts LinkstoLowGrowthinMexico.InNo Growth without Equity? Inequality, Interests, and Competition

1004

Journal of Economic Literature, Vol. XLVIII (December 2010)

and Competition in Mexico, ed. Santiago Levy and Michael Walton, 144. Houndmills, U.K. and New York:PalgraveMacmillan;Washington,D.C.:World Bank. Lucas, Robert E. 1988. On the Mechanics of EconomicDevelopment.Journal of Monetary Economics,22(1):342. Lustig,Nora.1998.Mexico: The Remaking of an Economy, Second edition. Washington, D.C.: Brookings Institution. Manzano, Osmel, and Francisco Monaldi. 2008. The Political Economy of Oil Production in Latin America.Economia: Journal of the Latin American and Caribbean Economic Association, 9(1): 5998, 10203. McKenzie,David,andHillelRapoport.Forthcoming. Can Migration Reduce Educational Attainment? Evidence from Mexico. Journal of Population Economics. McKenzie, David, and Christopher Woodruff. 2008. Symposium on Access to Finance: Experimental Evidence on Returns to Capital and Access to FinanceinMexico.World Bank Economic Review, 22(3):45782. Melitz,MarcJ.2003.TheImpactofTradeonIntraindustry Reallocations and Aggregate Industry Productivity.Econometrica,71(6):16951725. Mishra,Prachi.2007.EmigrationandWagesinSource Countries:EvidencefromMexico.Journal of Development Economics,82(1):18099. Moreno-Brid,JuanCarlos,JesusSantamaria,andJuan Carlos Rivas Valdivia. 2005. Industrialization and Economic Growth in Mexico after NAFTA: The Road Travelled. Development and Change, 36(6): 10951119. Ortiz,Guillermo.2003.LatinAmericaandtheWashington Consensus: Overcoming Reform Fatigue. Finance and Development,40(3):1417. Pages,Carmen.2010.La Era de Productividad.Washington,D.C.:Inter-AmericanDevelopmentBank. Pritchett, Lant, and Martina Viarengo. 2009. Producing Superstars for the Economic Mundial: The Mexican Predicament with Quality of Education. Unpublished. Rajan,RaghuramG.,andLuigiZingales.1998.FinancialDependenceandGrowth.American Economic Review,88(3):55986. Sokoloff,KennethL.,andStanleyL.Engerman.2000. Institutions, Factor Endowments, and Paths of Development in the New World. Journal of Economic Perspectives,14(3):21732. Stiglitz,JosephE.2003.Globalization and Its Discontents.NewYorkandLondon:Norton. Tornell, Aaron, Frank Westermann, and Lorenza Martinez. 2004. NAFTA and Mexicos Less-thanStellarPerformance.NationalBureauofEconomic ResearchWorkingPaper10289. World Bank. 2010. World Development Indicators. http://data.worldbank.org /data-catalog /world- development-indicators.

in Mexico, ed. Santiago Levy and Michael Walton, 11156. Houndmills, U.K. and New York: Palgrave Macmillan;Washington,D.C.:WorldBank. Haber, Stephen H. 1991. Industrial Concentration and the Capital Markets: A Comparative Study of Brazil, Mexico, and the United States, 18301930. Journal of Economic History,51(3):55980. Haber, Stephen H. 2005. Why Institutions Matter: Banking and Economic Growth in Mexico. Unpublished. Haber, Stephen H. 2009. Why Banks Do Not Lend: TheMexicanFinancialSystem.InNo Growth without Equity? Inequality, Interests, and Competition in Mexico, ed. Santiago Levy and Michael Walton, 283320.Houndmills,U.K.andNewYork:Palgrave Macmillan;Washington,D.C.:WorldBank. Hanson, Gordon H. 2006. Illegal Migration from Mexico to the United States. Journal of Economic Literature,44(4):869924. Hanson,GordonH.2010.SourcesofExportGrowth inDevelopingCountries.Unpublished. Hanson, Gordon H., and Raymond Robertson. 2010. China and the Manufacturing Exports of Other Developing Countries. In Chinas Growing Role in World Trade,ed.RobertC.FeenstraandShangJinWei,13759.ChicagoandLondon:Universityof ChicagoPress. Hsieh,Chang-Tai,andPeterJ.Klenow.2009.MisallocationandManufacturingTFPinChinaandIndia. Quarterly Journal of Economics,124(4):140348. Hsieh, Chang-Tai, and Ralph Ossa. 2010. A Global ViewofProductivityGrowthinChina.Unpublished. Jones, Charles I. 1995. R&D-Based Models of Economic Growth. Journal of Political Economy, 103(4):75984. Jurez,Laura.2008.AreInformalWorkersCompensatedfortheLackofFringeBenefits?FreeHealth CareasanInstrumentforFormality.InstitutoTecnolgicoAutnomodeMxicoDiscussionPaper08-04. Kaestner, Robert, and Ofer Malamud. 2010. SelfSelection and International Migration: New EvidencefromMexico.NationalBureauofEconomic ResearchWorkingPaper15765. Knox, Melissa A. 2008. Health Insurance for All: Evaluation of Mexicos Seguro Popular Program. Unpublished. Kuczynski, Pedro-Pablo, and John Williamson, eds. 2003. After the Washington Consensus: Restarting Growth and Reform in Latin America.Washington, D.C.:InstituteforInternationalEconomics. La Porta, Rafael, and Andrei Shleifer. 2008. The Unofficial Economy and Economic Development. Brookings Papers on Economic Activity,1:275352. Levy,Santiago.2008.Good Intentions, Bad Outcomes: Social Policy, Informality, and Economic Growth in Mexico. Washington, D.C.: Brookings Institution Press. Levy, Santiago, and Michael Walton. 2009. Equity, Competition,andGrowthinMexico:AnOverview. InNo Growth without Equity? Inequality, Interests,